It's an important topic, but there is also a meta-narrative here that's worth examining.
Here's the abstract Dana cited [which is scheduled to appear in the Journal of Consumer Research]:
Cost of Being Poor: Retail Price and Consumer Price Search Differences across Inner-City and Suburban Neighborhoods
This research undertakes a carefully designed and detailed empirical study to gain insights into (1) the extent of price differentials between wealthy and poor neighborhoods; (2) what induces such differentials, especially the nature and intensity of competitive environments, including mass merchandisers like Wal-Mart; and (3) their relative impacts. It finds a price differential of about 10%–15% for everyday items. Even after controlling for store size and competition, prices are found to be 2%–5% higher in poor areas. It also finds that it is not the poverty level per se but access to cars that acts as a key determinant of consumers' price search patterns.
The full article is in a pay-per-view section, so--to be bluntly honest--I went searching to see if I could find any more of it for free. Sometimes, you'd be amazed. I was particularly interested that the key determinant of consumers' price search patterns was access to cars.
I didn't find a freebie, but I did find this abstract [also in a gated site] from a Harvard study published in 2000:
Abstract Despite earlier evidence to the contrary, recent inquiries appear to reach a consensus that the poor pay more for food. However, these studies utilize samples drawn on the basis of prior knowledge of unfair pricing strategies, proximity of volunteer surveyors, or selected by other non-random methods. This paper revisits the issue of price discrimination by analyzing price data collected using a stratified, random sample design to answer the question of whether prices are higher in poor, urban neighborhoods. Contrary to the recent literature, it is found that market prices in poor neighborhoods are not higher than those in more affluent areas.
I am not suggesting at all that this study refutes or even directly opposes the one that Dana found. I can't read either of them, so I have no idea. Either, both, or neither could have serious methodological deficiencies. Either, both, or neither might qualify some of the statistical data in ways that would not be obvious in the abstract. [Abstracts, not surprisingly, sometimes make claims that the full article does not back up.]
What bothers me here is that--on both sides of the argument--you will easily be able to find people citing one or the other of these posts as authoritative evidence that their political ideology is right.
And fewer than 1% of those people will have actually read either goddamn article
Fewer than that, most likely, will be equipped with the math to critique the models employed by either researcher.
Instead, pundits and politicians will cite whichever article they happen to agree with [or, should I say, whichever article happens to agree with their pre-existing position] and declare their position vindicated, then state it's now time to move on with policy changes as a result of this proof.
I suspect that both researchers [who probably know each other and critique each other's papers at conferences while discussing how their families are doing] would be generally appalled to know that the people touting their work not only haven't read it, but wouldn't understand it if they did.
Dana and I disagree vehemently on lots of issues. But what we do agree on (and have consistently agreed upon) is the need for data, data that is rigorously tested and analyzed. To be clear: I'm not criticizing him at all for bringing this up; I think it's an important public policy discussion.
What I am bemused, befuddled, and bef**kt about, however, is the fact that in this country when we argue about facts, we are in general no longer using real facts to fuel the arguments.
Academic knowledge, once it leaves that academy (and sometimes before) has now yielded almost completely to sectarian dissonance.
That scares the hell out of me.